Recombinant Protein in Severe Pediatric Meningococcal Sepsis

Abstract & Commentary

Synopsis: In this multicenter trial of recombinant bactericidal/permeability-increasing protein (rBPI21) as an adjunct to antibiotics in children with severe meningococcal disease, the agent appeared safe but its effects were unimpressive.

Source: Levin M, et al. Lancet 2000;356:961-967.

In this randomized, placebo-controlled, multicenter trial, Levin and associates from the rBPI21 Meningococcal Sepsis Study Group enrolled children with clinical evidence of severe meningococcal sepsis to assess the safety and efficacy of rBPI21. The trial was carried out at 22 centers in the United States and the United Kingdom. Patients were eligible if they were between 12 weeks and 18 years of age, had presented with a fever plus a petechial or purpuric rash suggestive of meningococcal disease, and had severe disease as defined by a Glasgow meningococcal sepsis prognostic score (GMSPS) of 8 or more. Patients were excluded if there were signs of imminent death such as meeting clinical criteria for suspected brain death, ineffective ongoing CPR, "wide open" vasopressor therapy with ongoing hemodynamic deterioration, or clearly equivalent evidence of imminent death. Prospectively defined end points included mortality, mortality and amputations, amputations among survivors, and a change in Pediatric Overall Performance Category (POPC) scale to assess change in overall function at 60 days. Secondary end points included length of stay in ICU and in hospital, time on ventilator, and use of blood products.

Of the 1287 children evaluated, 395 were enrolled and 393 completed the study. Two patients (1 in each group) died after enrollment but before the study drug was infused. The treatment and placebo groups were comparable for various demographic parameters and for markers of inflammation. Overall mortality was 8.7% (34/393). The rBPI21 group had similar mortality (7.4%, 14/190) compared to the placebo group (9.9%, 20/203) (P = 0.48). Thus, the study failed to show a statistically or clinically significant reduction in mortality. Levin et al suggest that one of the problems could have been the lower than expected mortality in the placebo group. The rBPI21 group had a lower rate of "severe" amputation as defined by below knee or below elbow amputations. Treatment with rBPI21 also resulted in better POPC scale values when values were compared for each patient at their baseline. There was a reduction in blood product use and there were trends toward decreased ventilator duration and decreased length of ICU and hospital stay in the treatment group.


Sepsis and sepsis-induced multiorgan failure syndrome continue to be the major cause of mortality and morbidity in ICU practice. This somewhat unique trial looks at pediatric patients with meningococcal sepsis and takes a novel approach of using rBPI21, part of a naturally occurring protein. rBPI21 was safe in the study population but failed to improve mortality significantly. Bactericidal/permeability increasing protein is normally stored in neutrophils and has bactericidal properties. Thus, it acts as an adjunct to antibiotics to clear the bacteria.

As Levin et al have suggested, it is possible that rBPI21 may show statistical and, perhaps, clinically significant reductions in mortality if the control group has mortality in the expected range of about 25%. The problem may be how to predict that patient population. It is also unclear why the difference was significant for severe amputations only and not in the moderate or mild amputation group, as one would have expected that milder disease may be more responsive.

Increasing laboratory evidence suggests that outcomes in gram-positive infections are likely to be different when compared to outcomes in gram-negative infections. There is no specific therapy for sepsis-induced multiorgan failure. Anti-inflammatory approaches that target host inflammatory response have been repeatedly shown to be unsuccessful in improving the outcomes in several large trials. One of the problems with these studies is the heterogeneity of the population. It is also important to note that most of the agents in animal studies are used before or soon after the infection, where as in clinical practice there is a significant time lag between onset of infection and clinical manifestation of sepsis. The cause of multi- organ failure in sepsis also remains elusive. Early and appropriate antibiotic therapy will remain the mainstay of therapy for severe meningococcal disease until a subgroup of patients is identified that would most likely benefit from the addition of adjunctive therapy with rBPI21. (Dr. Nanavaty is Senior Research Fellow, Department of Critical Care Medicine, National Institutes of Health, Bethesda, Maryland.)